Post Thread

University at Buffalo philosophers Barry Smith and John Beverley explain that AI will not take human jobs because it cannot replicate the many forms of intelligence that lie outside its training data, and no system will ever reach true artificial general intelligence. Leading scaling approaches remain deeply flawed—enormously expensive and built on data tainted by human error—producing models that often undermine their own instructions while racking up unsustainable losses that point to a coming bubble. AI will keep delivering useful tools, yet the real story is how workers and students can prepare for the jobs it actually creates rather than the ones it is hyped to eliminate.

https://www.buffalo.edu/news/releases/2026/10/Why-AI-will-not-steal-your-job.html